143174 research outputs found
Sort by
Compliant piezoelectric medical micro-robots with visual servoing
No full textPrecision motion actuation is critical for microsurgery and catheter-based optical fibre diagnostic and intervention tools. Traditional actuation mechanisms, such as motor, magnetic, or pneumatic systems, are difficult to miniaturise while maintaining high-resolution motion. Piezoelectric materials, when used with compliant mechanisms, offer scalable, precise, fast, and high-force actuation with adequate motion range. This thesis explores an optical fibre steering technology suitable for catheters, diagnostic optical fibres, and microsurgical tool manipulation. The proposed technology combines piezoelectric beams with compliant motion translation structures, which have been validated through a series of experiments. A flexure-based compliant mechanism was designed based on flexure characterisation results, with piezoelectric benders used to implement a three-degrees-of-freedom delta robot. The fabrication of the system utilises additive manufacturing and origami structuring techniques. Closed-loop control is achieved through a novel on-board visual feedback system. Unlike conventional optical feedback systems, this fully internal visual feedback design enhances system compactness while providing precise and reliable camera-to-marker geometry alignment. Using this approach, the compliant delta robot is demonstrated with a motion accuracy of 7.5 μm, a resolution of 10 μm, and a precision of 8 μm. The robot successfully follows a range of programmable trajectories under these specifications and compensates for external forces typically encountered during operation. The integration of piezoelectric actuation, compliant motion translation, and onboard visual feedback is expected to deliver highly precise and reliable motion control for microsurgical applications.Open Acces
Lipid regulation of glucagon-like peptide-1 receptor in pancreatic beta-cells
No full textGlucagon-like peptide-1 receptor (GLP-1R), a key pharmacological target for treatment of type 2 diabetes and obesity, is a class B1 G protein-coupled receptor involved in the control of appetite and blood glucose levels via potentiation of insulin secretion from pancreatic beta-cells. Previously, cholesterol extraction from pancreatic beta-cells disrupted GLP-1R internalisation, clustering and cAMP responses. Therefore, the functional effects of the potential interactions between GLP-1R and its lipid microenvironment was studied.
This study highlighted that changes in the synthesis of complex sphingolipids using sphingomyelinase or eliglustat, caused changes in GLP-1R internalisation and a tendency for reduced cAMP signalling in pancreatic beta-cells. Reduced cholesterol synthesis using simvastatin increased GLP-1R-dependent cAMP signalling and insulin secretion in pancreatic islets. Increased cholesterol diet caused a decrease in the glucoregulatory effect of GLP-1R in vivo and reduced cAMP responses from extracted islets. Increased cholesterol levels ex vivo caused a reduction in cAMP signalling and insulin secretion from pancreatic islets.
This study identified potential sites/regions with high occupancy and residence times for cholesterol on GLP-1R using coarse grained molecular dynamic simulations. A screening of some key residues selected from these sites and detailed analyses of the effects of mutating one of these residues, valine 229 to alanine (V229A), highlighted the effects of changes in cholesterol binding to GLP-1R on receptor function. V229A mutant receptor caused changes in GLP-1R cholesterol interaction, plasma membrane activity, clustering, trafficking and signalling, causing improved insulin secretion in pancreatic beta-cells and islets.
In conclusion, this study highlighted the role of cholesterol in regulating GLP-1R function, identified relevant cholesterol binding sites on the receptor, and validated the effects of changes in cholesterol binding on receptor function. The results highlight the potential of GLP-1R cholesterol binding sites as locations that can be targeted for the rational design of novel allosteric modulators to fine-tune GLP-1R responses.Open Acces
Studies in branching processes - mathematical modelling of structured populations with applications to epidemics
No full textThis thesis treats of various types of branching processes, with some applications to epidemiology. Chapters 1 and 2 introduce a novel time-varying Crump-Mode-Jagers (CMJ) branching process whose governing integral equations are then relied upon in order to provide a rigorous mathematical justification to the renewal incidence equation in infectious disease modelling. Chapter 3 focuses on the asymptotic behaviour of a structured branching population where each individual in the population is characterised by a trait or a type whose dynamics follow a Markov process. The branching process is then expected to be driven by a positive triplet of first eigenvalue problem of the first moment semigroup. A strong law of large numbers for super-critical branching Markov processes is proven assuming convergence of the renormalized semigroup in weighted total variation norm. Convergence is obtained under an condition which provides a new Kesten-Stigum result in infinite dimension and relaxes the uniform convergence assumption of the renormalized first moment semigroup required in the work of Asmussen and Hering in 1976. Finally Chapter 4 treats of a novel variational auto-encoder, called -VAE, which has the property of being a proper stochastic process. As the combination of a generative model and a stochastic process that can be used as a prior, -VAE is then used to do full Bayesian inference. We provide particular examples with geo-spatial or epidemiological data.Open Acces
On explaining quantitative bipolar argumentation frameworks
No full textArgumentative explainable AI (XAI) has been advocated by several in recent years, with an increasing interest on explaining the reasoning outcomes of Argumentation Frameworks (AFs). While there is a considerable body of research on qualitatively explaining the reasoning outcomes of AFs with debates, disputes, or dialogues, explaining the quantitative reasoning outcomes of Quantitative Bipolar AFs (QBAFs) under gradual semantics has not received much attention, despite the widespread use of QBAFs in applications, such as voting polls and fraud detection, where explainability is crucial for ensuring comprehensibility and trust.
In this thesis, we contribute to filling this gap by proposing three novel theories of explanations. The first theory, Argument Attribution Explanations (AAEs), incorporates the spirit of feature attribution from machine learning in the context of QBAFs: whereas feature attributions identify the influence of features towards outputs of machine learning models, AAEs identify the influence of arguments towards a specific topic argument of interest. The second theory, Relation Attribution Explanations (RAEs), shifts the focus from arguments to the relations between arguments. RAEs measure the influence of relations towards a topic argument, providing a more fine-grained understanding than AAEs. The third theory is Counterfactual Explanations (CEs). Unlike AAEs and RAEs which explain the existing outcome, CEs suggest how to change the outcome to a desired one by modifying the QBAF in a cost-effective manner.
To evaluate these explanations, we theoretically study the desirable properties of the three proposed theories of explanations, including some new ones and some partially adapted from the literature to our setting. Additionally, we empirically validate the performance of these explanation methods, focusing on aspects such as scalability and robustness. Finally, we demonstrate the applicability of our proposed explanations by carrying out several case studies in various scenarios, such as fake news detection and movie recommendation.Open Acces
Expert opinion on the integration of combination therapy into the treatment algorithm for the management of dyslipidaemia: the integration of ezetimibe and bempedoic acid may enhance goal attainment
No full textThe clinically important link between LDL cholesterol (LDL - C) lowering and cardiovascular (CV) risk reduction is well-established and reflected in the 2019 European Society of Cardiology/European Atherosclerosis Society guidelines for the management of dyslipidaemia. They recommend a stepwise approach to reaching LDL - C goals, beginning with statin monotherapy at the highest tolerated dose. However, real-world data show a large gap between guideline LDL - C goal recommendations and their achievement in clinical practice. The treatment paradigm should shift from the concept of high-intensity statins to that of high-intensity, lipid-lowering therapy (LLT), preferably as upfront combination LLT, to overcome the residual CV risk associated with inadequate lipid management.
A multidisciplinary expert panel convened to propose treatment algorithms to support this treatment approach in patients at high and very high CV risk. The experts completed a questionnaire on the benefits of combination therapy and the role that novel LLTs, including bempedoic acid, might play in future guidelines. The integration of new LLTs into the suggested treatment algorithms for patients at high CV risk, very high CV risk, and those with complete or partial statin intolerance was discussed. Each algorithm considers baseline CV risk and LDL - C levels when recommending the initial treatment strategy. This expert consensus endorses the use of statin combination therapy as first-line therapy in patients at high and very high CV risk, and, in some circumstances, in patients with statin intolerance when appropriate. Given recent, compelling evidence, including real-world data, combination therapy as first-line treatment should be considered to help patients achieve their LDL - C goals
Hidden conflicts in neural networks and their implications for explainability
No full textArtificial Neural Networks (ANNs) often represent conflicts between features, arising naturally during training as the network learns to integrate diverse and potentially disagreeing inputs to better predict the target variable. Despite their relevance to the “reasoning” processes of these models, the properties and implications of conflicts for understanding and explaining ANNs remain underexplored. In this paper, we develop a rigorous theory of conflicts in ANNs and demonstrate their impact on ANN explainability through two case studies. In the first case study, we use our theory of conflicts to inspire the design of a novel feature attribution method, which we call Conflict-Aware Feature-wise Explanations (CAFE). CAFE separates the positive and negative influences of features and biases, enabling more faithful explanations for models applied to tabular data. In the second case study, we take preliminary steps towards understanding the role of conflicts in out-of-distribution (OOD) scenarios. Through our experiments, we identify potentially useful connections between model conflicts and different kinds of distributional shifts in tabular and image data. Overall, our findings demonstrate the importance of accounting for conflicts in the development of more reliable explanation methods for AI systems, which are crucial for the beneficial use of these systems in the society
Overview and outlook of thermal processes in geothermal energy extraction
No full textThis vision article accompanies a Special Issue of Applied Thermal Engineering dedicated to Heat Transfer in Geothermal Energy Extraction. This issue contains original research articles selected for publication in Applied Thermal Engineering, all connected by a focus on processes, technologies and systems for the exploitation of geothermal energy. Geothermal resources, which can be found at depths ranging from a few m to several km below the earth’s surface, are amongst the most promising renewable, zero carbon, and clean energy sources. Geothermal energy utilization technologies can be mainly classed into four categories: ground source heat pumps, energy geo-structures, enhanced geothermal systems, and closed-loop geothermal systems. Ground source heat pumps are the most well-developed geothermal technology for shallow depths, and have been widely used for building cooling and heating applications. Current and future heat pump research is focused on the coupling and hybridization of different heat pump configurations, using supplementary energy sources and phase change materials as storage media to overcome the thermal imbalance of the ground. Energy geo-structures, including energy piles, energy walls and energy tunnels, are an emerging and promising geothermal technology, with current research mainly focusing on simulating the effects of influencing factors on heat extraction effectiveness. Enhanced geothermal systems are a technology that can be used to extract deep geothermal energy for power generation and direct use. This technology has attracted significant renewed attention in the last decade. Future research should focus on the heat transfer mechanisms in high temperature rock fractures, and how the layout of inlet and outlet wells and different heat transfer fluids – such as CO2 – affect performance. Closed-loop geothermal systems, including coaxial and U-shaped systems, have no direct physical contact between working fluid and the reservoir, and may overcome some shortcomings of enhanced geothermal systems, such as leakage of the working fluid and water–rock interactions; however, further research is urgently needed on understanding the thermal recovery performance of these systems
Investigation of planar anisotropy evolution in aluminium alloy sheets under hot stamping conditions using digital image correlation
No full textHot stamping of aluminium alloy sheets is widely used for manufacturing high performance panel components across various industries. However, the anisotropic characteristics of the alloy and their evolution during deformation under hot stamping conditions, remain poorly understood, resulting in significant challenges in accurately determining its thermomechanical behaviour and developing
predictive models. To address this knowledge gap, a series of uniaxial tensile tests on a 1.5 mm thick AA6082 sheet under hot deformation conditions were conducted in this study using a Gleeble simulator at temperatures ranging from 350 °C to 500 °C and strain rates of 0.1 s-1 and 0.5 s-1. The planar anisotropy along both the length and width directions of AA6082 samples, as well as their
evolution during hot deformation was investigated by calculating the r-value (known as the Lankford (coefficient) based on the full-field strain distribution within the gauge length, measured using digital image correlation (DIC). The effects of strain fields selected from different regions within the gauge area on the calculated r-value were analysed. An empirical equation for r-value was proposed, for the first time, to model the planar anisotropy evolution across various deformation temperatures and strain rates under hot stamping conditions. This equation was subsequently applied to correct the stress-strain curves obtained using the C-gauge, an alternative strain measurement method, and the corrected data were compared with curves measured by DIC. This study provides insights on accurately determining thermomechanical behaviour and developing predictive models of aluminium alloys under hot stamping conditions
Multiscale modelling of charge and energy transfer in molecular solar cells and photosystems
No full textCharge photogeneration in natural and artificial organic photosystems share similar principles, involving light absorption, exciton diffusion, charge transfer, separation, and recombination, which collectively determine light-to-energy conversion efficiency. However, structural differences between the two systems lead to distinct limitations. In organic solar cells, most losses arise from non-radiative charge recombination at large donor-acceptor interfaces. In contrast, natural photosystems confine charge separation to a small reaction centre with only a few pigments. This structure reduces non-radiative losses but introduces limitations in the kinetics of charge transfer, particularly under high light intensities.
In this thesis, we study the charge separation process in both types of systems by a unified theoretical model, attending to their molecular structure. We coarse-grain the system to build a tight-binding Hamiltonian considering the molecular interactions. This approach provides a description of the system’s excited states, accounting for extended excitons and charge transfer states. By analysing the evolution of these excited states, we evaluate the yield of excitation recombination and energy storage, enabling to quantify the energy conversion efficiency of the system as a function of its structural and chemical properties.
Overall, this thesis presents a model to explore the correlation between the chemical and microstructural properties of molecular energy converters and their efficiency and losses. This work advances our understanding of how molecular structure and interactions shape the performance of molecular energy conversion systems.Open Acces
Search for Hidden Valley dark showers with displaced muons with the CMS experiment
No full textA search for signatures of a dark analog to quantum chromodynamics is performed. The analysis targets long-lived dark mesons that decay into Standard Model particles with a high branching fraction to muons. It is the first search at the Large Hadron Collider (LHC) that targets the decay of Hidden Valley dark showers into muons. A unique dataset with 10^10 B meson events is used. It was collected by the CMS experiment at the CERN LHC in 2018 using displaced muon triggers, which have high efficiency for the signal models. Resonant dimuon signatures are searched for, with both pointing and non-pointing topologies. No significant excess is observed beyond the Standard Model expectation. Upper limits on the branching ratio of the Higgs boson decays to dark partons are determined to be as low as about 10^-4, at 95% confidence level, surpassing and extending existing limits for the mean proper lifetime of less than approximately 0.1 m and for a mass as low as 2 GeV. First limits are set for extended dark shower models, probing the low-mass region down to 0.33 GeV.Open Acces